Building a cancer genetics and prevention program

Gastroenterologists offer more than just high-quality colonoscopy for colon cancer prevention. We often are the specialists who first recognize a genetic cancer syndrome during our endoscopy or clinic sessions. The patient who piqued my interest in colon cancer genetics was a 24-year-old woman who was referred for postoperative nausea after a hysterectomy for early stage uterine cancer (that alone should have raised alarm bells). Endoscopy revealed (by happenstance) a stomach coated with polyps. This led to a colonoscopy and diagnosis of familial adenomatous polyposis (uterine cancer within FAP is unusual but reported, for those of you studying for boards). In 1991, no coordinated genetics program existed within my practice so I arranged referrals to genetic counselors, surgeons, and pathologists. This led to the discovery of FAP and early stage (and curable) cancers in her two brothers and her father, in addition to extended pedigree analysis that established multi-organ cancer risks in other relatives. Years later, she brought her two adopted children to meet me and told me of lighting candles in my honor during an American Cancer Society walk. This is why we become doctors.

In this column, Dr. Xavier Llor describes the cancer genetics program he and others have built at Yale. It provides practical steps that can be taken by health system or community-based gastroenterologists to recognize and manage these complex syndromes. We are the specialists on the front lines and Dr. Llor helps us provide the coordinated care our patients expect from us.

John I. Allen, MD, MBA, AGAF
Editor in Chief

Among all common cancers, breast and colon have the highest percentage of cases that are due to hereditary syndromes. Many of the responsible genes have been identified, and the last few years have seen an increase in uptake of genetic testing supported by the refinement of the clinical criteria suggestive of these syndromes as well as the clear improvement in outcomes as a result of the adoption of cancer preventive measures in mutation carriers.¹ In spite of this, genetic testing for colorectal cancer (CRC) syndromes is not ordered as often as it should be according to the prevalence of these syndromes.² In contrast, testing for hereditary breast cancer has become more generalized, and the threshold for ordering genetic testing in the latter is often lower than for CRC. The are several reasons for this: 1) much greater awareness, by both providers and the general public, of hereditary breast cancer conditions; 2) fewer providers with expertise in CRC genetics; 3) lack of a systematic approach to identify patients with potential CRC syndromes; and 4) absence of a clear premorbid phenotype for the most common of all CRC syndromes, Lynch syndrome.³

The recent recommendation in practice guidelines to screen all CRC tumors for Lynch syndrome either with immunohistochemistry to evaluate mismatch repair (MMR) protein expression or through tandem repeat analysis to test for microsatellite instability⁴ has highlighted that about 10% of all CRCs (a percentage consistently seen in different ethnic groups⁵) need further cancer genetic evaluation, and many will require sequencing of germline DNA. Although data on cost-effectiveness of this approach are somewhat conflicting,^6,7 it is sensible because it is systematic, and studies have shown an increase in diagnostic yield through universal tumor screening.⁸ Unfortunately, in practice, often suspicious tumor testing results are not followed up by cancer genetics referrals, and many patients with CRC syndrome remain undiagnosed.

Patients with oligopolyposis (fewer than 100 polyps over time) also present diagnostic challenges. Some have attenuated familial adenomatous polyposis because of an APC mutation or MUTYH-associated polyposis. Recent findings have revealed other less commonly mutated genes that also result in oligopolyposis and a significant CRC risk: polymerases POLE and POLD1, GREM1, MCM9, or NTHL1. Because of the relatively low number of polyps in many of these syndromes and the lack of a systematic strategy to add up all polyps diagnosed over time, we not uncommonly fail to suspect some polyposis syndromes. Furthermore, the mixed pattern of polyps that is often associated with some of the mentioned mutated genes adds an extra challenge to diagnosing these cases.

In summary, there is a need to find systematic ways to triage and appropriately refer patients with a potential CRC syndrome to cancer genetics specialists so patients and their families can benefit from proper diagnosis and cancer preventive measures.